Learning Forum

[Ifexco]

Hi everyone,

I try to use the inbuilt element birth and death functionality in Ansys mechanical enterprise to kill the elements of two different bodies in two timestep operation but couldn't. It seems not more than one analysis (either to kill the element or make them alive) can be performed on the current time step. When I try to include another element birth and death functionality, it will capture the same time step that was put in earlier, and hence wouldn't run on a second time step. Please see figures below to illustrate this better.

I try performing element death on the highlighted portion in the first time step as shown below. But when I try performing another element death operation on the next portion (underneath) and changed the current step to 2, the first one I previously set was also changed to 2. As such, I am not able to perform separate death operation on the two time steps for the two portions. Can you please suggest what I can do to enable me perform this operation for these portions using the two different time-steps.

Thank you.

Regards,

Emmanuel

[Emperor]

HI,
In this case create another case and you make the dead element live in the 1st step.
Regards

[Ifexco]

Thanks @briveal. Yes, I have done so as show below in the figure. But I still need to kill the elements in the subsequent body (i.e., the body underneath the red highlighted one) using the 2nd step. I tried creating a new case for it but didn't run successfully. My question is how can I perform the element death of these two separate bodies using two different current steps?

[Emperor]

in the geometry part of detail of "element Birth and Death 3" you can select the number of bodies that will die. In your case there is only one body selected

[Ifexco]

,yes, I do know about that but in this case, I need to kill the elements in the bodies (one at a time). This is to say I need to kill the elements in body 1 in a time step 1 and then kill the elements in body 2 in type step 2. The problem is when I switch the current step to 2, all the other previously set cases also changed to step 2. When I run the simulation, it fails. Please what will you suggest I do in this case?

[Emperor]

What error do you get when the simulation fails?

[Emperor]

If I understand correctly:

step 1: death of element 1

step 2: death of element 2 below element 1 (element 1 still dead)



Is that right?

[Ifexco]

absolutely yes, you are right!

[Ifexco]

It was convergence error due to element distortion. But when I ran them separately on two different models, there was no error. I just think I am not yet getting the set-up right.

[Emperor]

in step 2, you have to select both bodies (dead), because if you only select body 2 in step 2, it will bring body 1 back to life I think and in my opinion you will have boundary condition problems.

[Ifexco]

,I have done this, yet still had the errors after running it. What more can I do?

[Emperor]

on this forum there are several topics that show how to debug models. First of all we need to know what type of error we had for that we need to search in the output of the solver the type of error (CRTL+F then 'error'), then according to the type of error we can bring corrective actions.

Have you made a good mesh ? Do you have non-linearities (contact, materials, geometry)?Have you applied the loading with the right order of magnitude to avoid instabilities? I will ask myself all these questions

A very interesting tip is to plot the Newton-Raphson residuals, generally they help to understand where the error comes from.

you can share your WB file so that I or someone from the community can take a look.

regards

[Ifexco]

thank you for your kind feedback. I have tried as much as I can to debug the model but to no success. Please find attached the WB file for the failed model. You recall, I had mentioned that the model ran successfully when I only used one step to kill the elements in the two bodies. I have also attached this model for comparison.
I await your kind feedback or from anyone else who can assist.
Thank you.